Immigration, intermarriage and the changing face of Europe in the post war period

(1)

post war period

Lucassen, Leo

Citation

Lucassen, L. (2009). Immigration, intermarriage and the changing face of Europe in the post war period. History Of The Family, 14(1), 52-68. Retrieved from

https://hdl.handle.net/1887/15000

Version: Not Applicable (or Unknown)

License: Leiden University Non-exclusive license Downloaded from: https://hdl.handle.net/1887/15000

Note: To cite this publication please use the final published version (if applicable).

(2)

Immigration, intermarriage and the changing face of Europe in the post war period

Leo Lucassen ^⁎ , Charlotte Laarman

Leiden University, The Netherlands

Abstract

In this article we compare the propensity to intermarry of various migrant groups and their children who settled in Germany, France, England, Belgium and the Netherlands in the post-war period, using a wide range of available statistical data. We try to explain different intermarriage patterns within the framework of Alba and Nee's assimilation theory and pay special attention to the role of religion, colour and colonial background. We therefore compare colonial with non colonial migrants and within these categories between groups with

‘European’ (Christian) and non-European (Islam, Hinduism) religions. First of all, religion appears to be an important variable. Migrants whose faith has no tradition in Western Europe intermarry at a much lower rate than those whose religious backgrounds correspond with those that are common in the country of settlement. The rate of ethnic endogamous marriages in Western Europe are highest in Hindu and Muslim communities, often regardless if they came as guest workers or colonial migrants. Whereas differences in religion diminish the propensity to intermarry, colour or‘racial’ differences on the other hand seem to be less important. This is largely explained by the premigration socialisation. Furthermore, the paper argues that the attention to institutions, as rightly advocated by Richard Alba and Victor Nee, needs a more refined and layered elaboration. Institutions, often as barriers to intermarriage, do not only emanate from the receiving society, but also—be it less formalized—within migrant communities. Especially religions and family systems, but also organized nationalist feelings, can have a profound influence on how migrants think about endogamy. Finally, strong pressures to assimilate, often through institutionalized forms of discrimination and stigmatization, not only produce isolation and frustrate assimilation (with resulting low intermarriage rates), but can also stimulate assimilation by 'passing' mechanisms. These factors, together with a more comparative perspective, are not completely ignored in the new assimilation theory, but—as this study of Western European intermarriage patterns stresses—deserve to be included more systematically in historical and social scientist analyses.

Keywords: Migration; Immigration; Intermarriage; Mixed marriages; Netherlands; Belgium; France; United Kingdom; Germany; Ethnicity; Race;

Religion; Gender; Islam; Muslims; Italians; Moroccans; Turks; Surinamese; Indians; Bangladeshi; Pakistani; West Indians; Algerians; Portuguese;

Spaniards; Greeks; Yugoslavians; Colonial migration; Colonialism; Guest workers; Labour migration

1. Introduction

Current public debates about immigration and integration in Western Europe focus strongly on what many perceive as the failing integration of non-Western

migrants, especially Muslims among them. The most pessimistic scenarios predict an ongoing cultural clash between‘them’ and the indigenous European population, point at the unwillingness of migrants and their offspring to assimilate, and fear the emergence of ethnic ghettos or

‘parallel societies’ (Huntington, 2004; Tibi, 2002).

Although it is clear that integration, both in the structural (work, education, housing) and the identificational

History of the Family 14 (2009) 52–68

⁎ Corresponding author.

E-mail address:l.a.c.j.lucassen@hum.leidenuniv.nl(L. Lucassen).

doi:10.1016/j.hisfam.2008.12.001

(3)

(marriage, friendship, associations) domain, is a slow and partly discordant process, many characteristics of the present situation are less new than is often assumed. Not only did European states experience large scale migrations before, but the paths of integration of these earlier migrants were also spangled with obstacles (Bade, Emmer, Lucassen, & Oltmer, 2007; Lucassen, 2005a).

This does not imply, however, that there are no differences with the past. In contrast to the post war period, before 1940 the proportion of non-European migrants was insignificant and the integration of some groups seemed to evolve slower and with more difficulties than in previous periods (Lucassen, Feldman, & Oltmer, 2006). This becomes apparent both in poor school results and high drop out rates and disproportionate high unemployment figures among the children of immigrants. Think of the offspring of Algerians in France, Moroccans and Turks in the Netherlands, Belgium and Germany and Pakistani and Bangladeshi in the United Kingdom (Crul & Vermeulen, 2003; Thomson & Crul, 2007). Moreover, part of the immigrants with an Islamic background seems to oppose the core values of Western European societies, such as equality between men and women, freedom of speech and the separation of church and state.

Although there are clear signs that a European moderate form of Islam is developing (Klausen, 2005;

Laurence & Vaisse, 2006; Modood, 2007; Roy, 2004), the cultural distance between the offspring of Muslim migrants and the indigenous population has not disappeared and makes itself felt in the identificational domain. In this paper we concentrate on one of the most salient aspects of integration, intermarriage, as expressed by the propensity of immigrants and their descendants to marry with partners from the native population. The rate of intermarriage is especially interesting because it tells us something about the social and cultural distance between immigrant groups and the native population. It is even more telling than other indicators of integration, as the choice of a life partner has a pervasive influence

on the day-to-day life. Intermarriage is also linked to the structural realm, however, because in general upward social mobility is coupled with increasing rates of mixed marriages (Lievens, 1998; Mutarrak, 2003; Qian, 1997). Therefore intermarriage can also be regarded as an indicator of the structural side of the integration process (Alba & Nee 2003; A. Gordon, 1964; Hwang, Saenz, & Aguirre, 1995). A good example is the position of West Indian immigrants in the United States whose intermarriage rates, compared to the African Americans who settled there much earlier on, are significantly higher. Furthermore, West Indians have higher educational qualifications and distinguish themselves from the African Americans by their British English which is associated with a higher class position (Foner, 1998).

In this article we compare the propensity to intermarry of various migrant groups and their children who settled in Germany, France, England, Belgium and the Netherlands in the post-war period (seeTable 1). We are interested in the relation between intermarriage and the overall integration process, with special interest in the role of religion and colonial background, factors that influence the propensity to intermarry (Kalmijn & Tubergen, 2006;

Lucassen, 2002c; van Niekerk, 2004). We therefore compare colonial with non colonial migrants and within these categories between groups with‘European’ (Chris- tian) and non-European (Islam, Hinduism) religions.

Because data on intermarriage are largely lacking for refugees, we have left this category out. The groups analysed in this overview are summarised in the next table.

Before we present the data on intermarriage patterns, we will first discuss briefly the relevance of intermarriage and its place in the current assimilation theory.

2. Intermarriage and assimilation theory

Marriage patterns have always been of great interest to migration scholars. In their recent study on assimilation

Table 1

Schematic overview of groups that are analysed

Netherlands Germany France Belgium UK

Tu Mo Wi In Tu It Gr Yu Al Sp Po Tu Mo Pa Ba Wi In

GW-NER X X X X X X

COL-NER X X X X X

GW-ER X X X X X

COL-ER X X

Key: GW = guest workers; COL = colonial migrants; ER = European religion (Catholic, protestant, Jewish); NER = Non European religion (Muslims, Hindus); Tu = Turks; Mo = Moroccan; Wi = West Indians (including Creole Surinamese); In; Indians (including Hindustan Surinamese); It = Italians;

Gr= Greeks; Yu = Yugoslavs; Al = Algerians; Sp = Spanish; Po = Portuguese; Pa = Pakistani; Ba = Bangladeshi.

N.B. As Algerians share elements of the colonial and the guest worker category, we have put them in both categories.

(4)

processes in the United States Richard Alba and Victor Nee posit mixed marriages as the ultimate litmus test for assimilation. In the tradition of the Chicago School of Sociology and followingMilton Gordon's (1964)seminal work on assimilation they argue that a high frequency of mixed marriages is in general a sign of decreasing social and cultural distance between ethnic groups, showing that social and cultural differences are not regarded anymore as an obstacle to marry out by both the minority and the majority group. As a result, ethnic or racial boundaries will blur or even fade away (Alba & Nee, 2003; M. Gordon, 1964, p. 205–206).

The relation between intermarriage and the fading of group boundaries is not unambiguous (Sinke, 2002, p. 25–26 and p. 235 footnote 62). There are also other factors that influence intermarriage, as the length of stay in the receiving country, the age and marital status at migration, and the idea to once return to the country of origin. For instance, the Spanish guest workers in the Netherlands had low rates of intermarriage; not due to a large cultural gap between the Spaniards and the Dutch, but because most of them were already married and returned (or strongly wished to return) to Spain after a couple of years. Taking these aspects into account, many scholars nevertheless more or less follow Alba and Nee's assumptions. Currently, the marriage behaviour of migrants with a low socio-economic position and a different cultural background, like those from Muslim countries (Turkey, North Africa, South East Asia), draws a lot of attention, both from researchers, policy makers and the media. The fact that children of migrants in Western Europe predominantly marry within their own group, often with someone from the country of birth of their parents, is considered a serious hurdle on the road to integration.¹The relation between structural and identificational integration is not always so straightforward, however. Indians in the United King- dom, for example do very well at school and in the labour market, in some respects even better than native English, but they overwhelmingly marry co-ethnics (Modood & Berthoud, 1997, p. 345). Nevertheless, Indian women with higher qualifications have a higher propensity to intermarry (Mutarrak, 2007, chapter 6), and economic analyses of current intermarriage patterns show that migrants who intermarry earn significantly higher incomes than endogamous married immigrants, even when we take account of human capital endow-

ments (Meng & Gregory, 2005; Dribe & Lundh, 2007; van Niekerk, 2002, p. 127;van Heelsum, 1997, p. 120–125).

Notwithstanding the relative unanimity among scholars with respect to the importance of marriage patterns for understanding and measuring assimilation, trends in marriage rates do not simply speak for themselves. First of all we have to define what we mean by a ‘group’. In most migration studies it is assumed that the most important criterion is origin, defined in territorial (state/region) terms. In this context ‘mixed’ refers to nationality and/or ethnicity.

Group ties, however, are not only—and not always primarily—determined by national and regional iden- tities. Until the 1960s, for most people religion was a greater barrier for marriage than nationality (Hondius, 1999, p. 136; Lucassen, 2005b).² Some American scholars in the 1940s therefore predicted assimilation along religious lines, creating multiple melting pots (Kennedy, 1944).

Whereas in secularized Western European societies, in contrast to the United States (Foner & Alba, 2008), religion has lost its salience in the second half of the 20th century, this is not the case for migrants coming from religious societies, especially Muslims and Hindus. In these communities, religion often remains highly relevant and can obstruct marriages with partners with no or a different religion. We should add that indigenous European men and women, religious or not, also have great hesitations to marry a Muslim (for Hindus this is probably less the case), so that the group boundary between Muslim migrants and others is double edged. Secondly, in some migrant communities (for example West Indians), cohabitation is often more common than marriage. In marriage statistics their intermarriage rates appear to be lower, whereas they have long lasting out-group relationships that in practice do not differ much from a formal marriage. We are aware of this bias in marriage statistics but there are no comparable data available on cohabitation rates of the different ethnic groups in France, the UK, the Nether- lands, Belgium and Germany.

An aspect that has not lost its salience is class.

Western European societies may have embraced merito- cratic ideals, it is clear, as Bourdieu,Wacquant (1996) and many others have argued, that class remains highly relevant, leading to widespread social endogamy, also known as class homogamy (marrying within the same class) (van Leeuwen & Maas, 2005, p. 1). Not only

1For France Tribalat (1995); for Germany Venema and Grimm (2002). For the United Kingdom,Dale and Holdsworth (1997) and Modood and Berthoud (1997); and for the NetherlandsHooghiemstra (2003, p. 3–4).

2With the immigration of Islamic migrants after World War II religion, linked to ethno-cultural perceptions, has again acquired a master status to use Hughes terminology:Hughes (1945).

(5)

because people are prejudiced, but simply because most people prefer a partner who shares one's values and tastes (van Leeuwen & Maas, 2005, p. 17; Kalmijn, 1994).

In this article we have chosen the prism of ethnicity, using intermarriage statistics structured around origin and differentiated for sex, age, generation and religion, but—alas—mostly lacking information about social class.

Given the close relation between one's social position and the likelihood to marry outside one's ethnic group, increased ethnic intermarriage rates at the group level may to some extent be interpreted as a proxy for upward social mobility (Lucassen, 2005a, p. 70). In other words we expect that intermarriage rates will rise with increasing structural integration. The exception to this rule are situations in which culturally and institutionally embedded discrimination cause and endorse social distance, as in the cases of African Americans (Foner & Fredrickson, 2004;

Model & Fisher, 2002), or when religious boundaries are drawn by groups themselves like orthodox Jews, Muslims and Hindus. Intermarriage patterns therefore not only shed light on the integration process, but also on the criteria that in certain historical contexts are deemed relevant, both by migrants and established populations.

2.1. Determinants of intermarriage

Scholars from different disciplines have formulated various theories to explain intermarriage (or the lack thereof), within which we can roughly distinguish two variants. First of all, the contact theory assumes that people have to meet before they can start a relationship.

This seems a truism, but the extent to which different groups (in our case migrants and natives) attend the same schools, live in the same neighbourhoods, work in the same places, go to the same clubs, bars or worship together, strongly influences the propensity to intermarry (Kalmijn, 1998). A similar argument can be found in Peter Blau's work who elaborated Simmel's inter- pretation of the modernization in the 19th century. In contrast to traditional societies in modern complex social orders, Blau argues that “a person's multiple group affiliations constitute mostly crosscutting circles”, that further profound and lasting intergroup relations. To what extent people from different groups do intersect depends, among other things, on sex ratios, age distribution, relative group size, and the heterogeneity of the group (Blau, Blum, & Schwartz, 1982; Blau, Beeker, & Fitzpatrick, 1984; Glick, 1976).

Unbalanced sex ratios often influence the propensity to intermarry. When migrant groups consist predominantly of men, the chances of marrying someone from

their own group diminish. Examples are single male guest workers from Italy, Spain and Portugal, but also from Turkey and Morocco in Western Europe during the 1960s, many of whom married native women.³For the Netherlands we have the examples of Italian chimney sweeps in the nineteenth century, and Chinese sailors in the 1930s and the 1940s (Chotkowski, 2006; Hondius, 2000).

The contact theory describes a situation in which serious social or institutional barriers are absent. In many cases, however, life is not that idyllic and this is where the barrier theory comes in, predicting low intermarriage rates when secular and/or religious authorities put up (institutional) barriers to restrict or discourage marriage across social, religious, racial or national lines. Examples are the anti-miscegenation laws in the United States, which were only abolished by the Supreme Court in 1967, migration laws that aim at restricting marriage migration or the refusal of priests to solemnize religiously mixed marriages. Barriers, however, do not only transpire from the dominant society, they can also emanate from (migrant) groups themselves, often linked to political, cultural or religious reasons (de Hart, 2006; Hondius, 2000; Kennedy, 1944; Moran, 2001). Sometimes in reaction to discrimination, as in the case of the Black Power movement in the U.S. who from the 1960s onwards rejected marriages with whites (Spickard, 1991), but groups such the Pennsylvanian Amish may also prefer endogamous marriages indepen- dent of exclusionary practices (Dorsten, Hotchkiss, &

King, 1996).

Both approaches have been incorporated in the modernized assimilation theory as developed by Alba and Nee. Their framework allows us to study in a systematic way the variables listed in both the contact and the barrier theory. When intermarriage rates are low, the latter allows us to focus on the attenuation of (perceived) cultural, socio-economic, ethnic or racial differences between groups and the effect on the propensity to intermarry. Crucial is the measure to which these differences are seen as important or as a problem (Merton, 1941). If the partners look upon each other as too different to share a life together or their family and friends react negatively upon perceived differences, the couple is likely to split up or will not start a relationship in the first place. On the other hand the contact theory is important in highlighting the actual chances of meeting and the existing social distance between groups.

3Glick (1976)called this the‘marriage squeeze’: Glick, American families. See also Tables 9 and 14 in the Appendix A.

(6)

Embedding contact and barrier notions in the assimilation theory has the advantage that intermarriage can be systematically linked to other dimensions and studied in a longer time frame. Thus it predicts that over time, often over generations, the descendants of migrants will intersect more with the native population and at the same time will overcome social pressure to choose marriage partners from within their own group. Only when institutional barriers stay in place, consolidating group boundaries, intermarriage rates are bound to remain low. In most historical cases that we know of, at least in Western nation states, this assimilation perspective works quite well, as is illustrated by the migration history of Western Europe before World War II.

3. Migration and intermarriage before World War II In contrast to the United States, most European states have considered themselves ethnically homogeneous from the 19th century onwards, or—as in the case of Germany and its Polish speaking minority— tried to homogenise their populations as much as possible (Weber, 1976). As far as the collective memory of these countries included migration, it was limited to the mass emigration to overseas destinations, especially the Americas. Thus, between 1840 and 1920 almost 55 million Europeans, first from the North-West and later also from the South and the East left the old continent to settle overseas, especially in North America, although considerable numbers returned (McKeown, 2004; Nugent, 1992; Wyman, 1993).

This stress on emigration has buttressed the nationalist idea that immigration to Western Europe was a rather unnatural and recent phenomenon. Since the 1980s, however, migration historians have successfully refuted this received knowledge (Bade, 2000; Hoerder &

Moch, 1996; Lucassen, 1987; Moch, 2003). We now know that many states experienced massive internal movements, mostly to industrial core areas, whereas millions of intra-European migrant crossed state borders. To mention just the most striking examples:

hundred thousands of Italians who settled in France, the equally large numbers of Irish who ended up in Lancashire, Glasgow and London, and Polish speaking workers who drastically changed the population of what was to become the Ruhr area (Lucassen, 2005a).

The importance of the intra-European migrations is well illustrated by Italian migrants in the 19th and 20th centuries, more than half of whom did not board ships for the new world but moved within Europe (Bertagna

& Maccari-Clayton, 2007; Gabaccia, 2000). In fact,

both internal and international migration were a structural feature of European societies at least since the early modern period, as a recent overview has convincingly shown (Bade et al., 2007; Canny, 1994).

The idea of stable and ethnically homogeneous populations, so dominant in national historiographies, is therefore highly problematic, although the impact of foreign immigrations differs from country to country (seeTable 2).⁴

Until the Second World War the bulk of these migrations concerned Europeans. Immigrants from non- European regions, also those who stayed only tempora- rily, were rare, especially in North-Western Europe. This would only slowly change after World War I. During the Great War France recruited more than two hundred thousand workers and soldiers from North Africa and Indochina. Although most returned to their country of origin, the migration paths remained intact. In France, ten thousands of Algerian and other North African men, who from 1914 onwards had more or less free access to the ‘Hexagone’, found employment as low skilled workers in the 1920s and 1930s (Lewis, 2007;

Rosenberg, 2006). In other countries the numbers of colonial migrants were much lower, but also in England and the Netherlands numbers went up.⁵Finally, Europe came into contact with Chinese migrants, mostly boilermen and sailors, but also traders, who constituted small but highly visible male communities in European harbour cities like Liverpool, Marseilles, Rotterdam and Amsterdam.

Table 2

Foreign (born) population in Western European countries (1850– 2000)

1850 1910 1930 1950 1970 1990 2000

UK^a 4 2 2 (?) 5 6 7 8

Germany 1 2 1 1 5 8 9

France 1 3 7 4 6 7 9

Netherlands^a 2 2 4 1 4 8 10

Belgium 1 4 4 4 7 9 9

Switzerland 3 15 9 6 17 18 21

The percentage of the Irish-born population in 1851 was 2.9% for England and Wales. In Scotland the impact was even greater (7.2%):

Lucassen (2005a, p. 31).

Source:Bade et al. (2007),Herbert (1990),Mauco (1933).

aForeign born. In case of the UK, we have included the Irish.

4The British Isles taken together, for example, have remained rather isolated already since the early modern period, with many people leaving, but relatively few entering until the mid 20th century:van Lottum (2007),Lunn (2007).

5This includes colonial migrants from mixed descent born in the colonies:Bosma (2007).

(7)

3.1. Intermarriage

Given the racially loaded negative stereotypes of non-Europeans it is surprising that many migrants who came to Western Europe between the wars, like Chinese and West Indians, married European women. This is partly explained by the one-sided sex ratio of these groups, that were almost entirely composed of young unmarried men. Especially a considerable part of the (tiny) Chinese male population married indigenous women, especially after they realised that returning to their home country was difficult, or even impossible (during World War II).⁶An exception to this pattern are Algerians and other North African French subjects in France who, although legally being citizens, were stigmatized, marginalised and discriminated against.

This resulted in systematic surveillance by the French police, regular expulsions to North Africa and curtailment of political and social rights (Lewis, 2007, p. 188–215;

Rosenberg, 2006). This severely reduced the possibilities and tendency to intermarry.

For intra-European migrants, like the Italians in France, Germans in the Netherlands and the Polish miners in Belgium, to mention a few well known groups, exogamy rates depended not only on their sex ratio, but also on religion, marital status of the migrants and the prominence of nationalistic feelings, for example among the Poles and the Irish (Belchem & Tenfelde, 2003;

Lucassen 2005a; Oenning, 1991). Whereas thousands of German female servants in the Netherlands married Dutch men (Lucassen, 2002a,b), Poles, with a more balanced sex ratio and a vigorous nationalism, married mainly within their own group (Beyers, 2007, p. 119;

Lucassen, 2005a, p. 96;Girard & Stoetzel, 1953, p. 70;

andMauco, 1932, p. 532). How their children fared is largely unknown. Except for the Poles in the Western part of Germany (with intermarriage rates up to 30%

around 1920) and the Italians in France (probably over 50%) (Lucassen, 2005a, p. 69–71), research is almost entirely lacking.

As the pre-war situation largely confirms the assumptions of the assimilation theory, it also reveals some weak or at least underdeveloped spots, that force us to look more closely at the role of institutional barriers. Alba and Nee, concentrating on European, Asian and Hispanic migrants to the United States,

predominantly point at the increase of intermarriage in the longer, intergenerational, run. Important factors in the blurring and fading of group boundaries are upward social mobility by the descendants of migrants and the decrease of ethnic stereotyping over time, especially from the 1960s onwards. The only exception being the boundary between black and white Americans.

Alba and Nee pay less attention to efforts to promote endogamy within ethnic groups,⁷ and do not explicitly link intra-group pressure to their institutional framework. The case of the Poles in Germany and the Irish in England before World War II, however, shows that it is important to broaden Alba and Nee’s definition of institutions and not restrict it to the legal structures of the receiving society, such as the anti-miscegenation laws in the Jim Crow era, nor to attitudes of the native population more in general. Especially the Polish story shows that the role of institutions is more complex. On the one hand, the militant Polish ethnocentrism in the period 1870–1930 was not only a reaction to the repressive policies of the German state, but was also a crucial part of the Poles' own national project.

On the other hand, the systematic repression of their nationalistic project by the German state in the longer run lead many of their children born in the Western part of Germany to pass as Germans. A clear sign is the increase in the number of applications to change their surnames, which shows that they succumbed to the direct and indirect assimilation pressure. With its stress on institutions, the new assimilation theory seems in the short run to under- estimate the influence of barriers within immigrant groups, whereas in the longer run it should pay more attention to ‘passing’ as a reaction to institutionalised assimilation pressure.⁸ In the next section we will use this enriched assimilation theory to analyse intermarriage patterns in post war period, concentrating on the marriage behaviour of the second generation at the end of the 20th century.

4. Data and method

Immigration to Western Europe, from the Southern fringe and from Asia, Africa and the Caribbean, increased since the late 1940s, caused by decoloniza- tion, guest worker programmes and—especially from the 1980s onwards—refugees from Asia and Africa.

Their integration process has been studied from many

6 There is no systematic study, let alone an international comparison, but many case studies give ample indications that mixed marriages were not uncommon:Lucassen (2005a, p. 123). See also van der Harst and Lucassen (1998, p. 136). For Algerians see Rosenberg (2006, p. 135–138).

7They only briefly refer to Jews (Alba & Nee, 2003, p. 92).

8This phenomenon is only mentioned once (Alba & Nee, 2003, p. 61).

(8)

angles, including intermarriage. To what extent those who settled in Western Europe did intermarry in the period 1950–1980, however, is largely untrodden terrain. Only in the last decades of the 20th century has information on intermarriage patterns of migrants and their children become available. From the 1980s onwards, national statistical bureaus have started measuring these ‘social facts’ and research institutes, like the French INSEE and the Dutch SCP, that became interested in this issue thereby made intermarriage

‘legible’ (Scott, 1998). But scholars have also gathered data through surveys and micro censuses. So far, however, a systematic comparison of intermarriage patterns throughout Western Europe is lacking.

Notwithstanding the increasing statistical apprehen- sion of mixed marriages of migrants and their children in Western Europe, it was not easy to find comparable and consistent data on intermarriage for the first and second generation of the groups we analyse in this paper. Partly because national traditions in registering ethnic difference differ, but also because definitions vary as well; from registrations on the basis of nationality (Germany, France) to origin and ethnicity (Netherlands and Great Britain). Finally, the periods under observation, and thereby the cohorts that are measured, fluctuate. Nevertheless, the available data on

intermarriage rates of first and second generation migrants in the period 1992–2003 reveal a number of interesting patterns, that are supported by secondary literature (Kalmijn & Tubergen, 2006; Neyrand &

M'Sili, 1997; van Niekerk, 2004, Tribalat, 1995). We summarize our results inTable 3.

The most important trend we discern in the data on intermarriage is that religion matters much more than colour or ‘race’, which is in stark contrast to the race obsessed American case. Migrants of African origin, but with a Christian religion, tend to intermarry far more than migrants with a lighter skin colour but with a non- Western European religion (Islam, Hinduism). This is true for both men and women, although the rates for women are even lower than for men, which is explained by the patriarchal traditions in Muslim and Hindu societies (Todd, 1985). Although such cultural norms are neither homogeneous nor static (Grillo, 2008), in the last decades their influence cannot be underestimated.

For example, Muslim women who marry a non-Muslim man are often seen as lost for the (patrilineal) family and thereby for Islam. From this perspective the children will take the religion of the non-Muslim father. This is different for Muslim men who marry a non-Muslim woman as men are not seen as lost for the family and faith and his children are expected to be Muslim (Todd,

Table 3

Average intermarriage rates in Western Europe 1990–2000 according to sex and generation

Male 1st Male 2nd Female 1st Female 2nd Average

COL-ER West-Indians (UK, NL) 26 60 26 46 40

GW-ER Italians, Greeks, Yugoslavians (Ger), Spanish, Portuguese (Fr.) 22 48 15 38 31

GW-NER Moroccans, Turks (Ger, NL, Be) Algerians (F.) 11 16 5 8 10

COL-NER Algerians, Moroccans (Fr.), British-Indians, Pakistanis, Bangladeshis (UK), Indo-Surinamese (NL)

5 11 7 10 8

Key: SeeTable 1. The rates are averages ofTables 4–7.

Table 4

Intermarriage rates for men of the first generation (1991–2003)

Netherlands (2003)^a Germany (2000) France (1992) Belgium (1991) UK (2002)

Tu Mo Wi^b In Tu It Gr Yu Al Sp Po Tu Mo Pa Ba Wi In

GW (NER) 7 6 13 15 7^c 17^c

COL (NER) 10 3 1 5

GW (ER) 38 18 21 18 15

COL (ER) 26 27

Key: SeeTable 1.

Source: see Appendix A, Table 1 (only the rates from 2000), 4 (1991), 5 (1988–2002: only for Indians in the Netherlands), 6 (2003), 12 (2002) and 13 (1992).

a Except for the Indian Surinamese in the Netherlands (1988–2002: see Table 5 in Appendix A).

b The statistics on the Surinamese do not differentiate between Creoles and Indians. We know however, that the rates for Indians are much lower (around 10%: see Table 5 in Appendix A), so that the average rate for all Surinamese (26) is severely debased.

c The Belgium rates do not distinguish between first and second generation, and are therefore most probably too high.

(9)

1994, p. 178 and 306). But this only applies when they marry a Christian or Jew (‘religions of the book’). It is striking that the intermarriage rates of children from parents with a non-European religion do rise somewhat, but on average remain low. This is true for Hindus as well as for Muslims (see Tables 4–7).

We should beware, however, to explain these differences strictly in religious terms. Religion is not an isolated characteristic, but closely linked to cultural traditions and family systems (Todd, 1985). Decisions about crossing ethno-religious boundaries are therefore not only motivated by theological, but also by cultural arguments, linked to (patriarchal) family systems and traditions. An outstanding feature of in-group marriages among Muslim migrants is that they almost without exception are made within their own ethnic group (Turks marry Turks, Moroccans marry Moroccans), showing that extra religious group factors play a role as well. The most important cultural factor seems to be the predomi- nant“endogamous, community-based family system”, as explained by the anthropologist Emmanuel Todd (build- ing on F. Le Play's mid 19th century typology), who

argued that in most Muslim societies it is acceptable to marry cousins, especially children of brothers, in order to maintain the unity of the patrilineal clan (Todd, 1994, p. 284–285). As we will see, Todd's stress on family systems can shed light on a number of deviations from the general dichotomy between religion and colour. Although differences in religion between migrants and nationals of the receiving countries are a strong predictor for low intermarriage rates, it has its limits, as the following examples will show.

5. Explaining intermarriage patterns in Western Europe since the 1950s

Among Portuguese women in France, who share their Catholic faith with the French nationals, a strong assortative mating pattern is visible, especially in the first generation (Table 5). The bulk of the Portuguese migrants arrived as guest workers in the period 1963– 1973. At first, they were mainly men, but they were very soon followed by women and children. This migration was so massive that around 1980 some

Table 6

Intermarriage rates for men of the second generation (1992–2003)

Netherlands (2003) Germany (2000) France (1992) UK (2002)

Tu Mo Wi In Tu It Gr Yu Al Sp Po Pa Ba Wi In

GW (NER) 7 14 13 30^a

COL (NER) 9 10 13

GW (ER) 51 35 30 66 59

COL (ER) 59^b 61

Key: SeeTable 1. Data for Belgium are not available for the second generation.

Source: see Appendix A, Table 1 (only the rates from 2000), 6 (2003), 12 (2002), 13 (1992).

aEstimate: percentage of all first unions (including cohabitations) is 50.

bAverage of Surinamese (43%) and Antilleans (75%).

Table 5

Intermarriage rates for women of the first generation (1991–2003)

Netherlands (2003)^a Germany (2000) France (1992) Belgium (1991) UK (2002)

Tu Mo Wi^b In Tu It Gr Yu Al Sp Po Tu Mo Pa Ba Wi In

GW (NER) 9 7 7 9 2^c 6^c

COL (NER) 10 2 1 5

GW (ER) 13 12 18 25 6

COL (ER) 35 16

Key: SeeTable 1.

Source: see Appendix A, Table 1 (only the rates from 2000), 4 (1991), 5 (1988–2002: only for Indians in the Netherlands), 6 (2003), 12 (2002) and 13 (1992).

a Except for the Indian Surinamese in the Netherlands (1988–2002: see Table 5 in Appendix A).

b See legend underTable 4(b).

c The Belgium rates do not distinguish between first and second generation and are therefore most probably too high.

(10)

630,000 Portuguese had settled. The intermarriage rates of women in this group, that is regarded as perfectly integrated in other spheres of French society, are significantly lower than of other Southern European and Catholic guest workers like the Spanish and the Italians. This might partly be explained by the fast follow migration and thereby the rebalancing of the sex ratios, but this in itself might as well be caused by a more general ethnic cohesiveness among the Portuguese.⁹ This is reflected in the close contacts they keep to their home country, the strong and dense associative networks (Simon, 2003, p. 1097), and low naturalization figures (Borrel & Tavan, 2003, p. 113 Simon, 2003, p. 1095).

Ethnic feelings among Portuguese were so strong that intermarriages among the first generation were frowned upon and mixed couples were often excluded from the ethnic community. According to Patrick Simon, this exclusion most probably has negatively influenced the societal chances of the children from these mixed marriages who on average do less well in school and in the job market (Simon, 2003, p.1106). In their case, the general rule that ethnic exogamy is related to upward social mobility is therefore not corroborated. This penalty on intermarriage, however, has decreased among the second generation, among whom the rates have risen spectacular (Tables 6 and 7).

The second divergent pattern is found among Algerians in France, who have a colonial background with elements of the guest worker recruitment system. Although being Muslims and coming from the same cultural North African (Maghreb) region, their intermarriage rates are significantly higher than those of Moroccans in the Netherlands and probably in Belgium as well.¹⁰Accord-

ing to Todd, the relatively high intermarriage rates among French Algerians can be explained by the colonial link between France and Algeria; bonds that are lacking in the case of Moroccans in Belgium and the Netherlands (Todd, 1994; p. 307;Shepard, 2006, p. 20–21). From the time that Algeria became a French département in 1848, Algeria underwent a strong Francophone influence;

linguistically, culturally, economically and politically (MacMaster, 1997; Lucassen, 2005a, p. 173–179).

Hundred thousands of French‘colons’ settled there and the French system of government, with départements, arrondissements, communes, prefects and maires was put in place. The result was that many Algerians who went to French were, just like other colonial migrants, already partly socialised in the French culture, including the ideology of egalitarian individualism.

It is remarkable that the widespread collective negative attitude, and institutionally embedded image of Algerians already before World War II did not refrain French and Algerians (men and women) to intermarry (Rosenberg, 2006, p. 166–167, 190). This does not mean however that this negative image and the partly traumatic colonial history did not influence intermarriage trends. Neyrand and M'Sili (1997) found that Moroccan and Tunisian migrants, also Muslims and colonial migrants, had higher rates of intermarriage, especially among the first generation, than Algerians.

They explain this by the animosity of the Algerians towards the French as a consequence of repercussions in the Algerian war of independence.

If we compare intermarriage patterns of Algerians and Turks, we find that in contrast to the Turks, among Algerians, the French socialisation and reduced social distance as a consequence of colonialism eroded the endogamous community family system. The strong influence of the cultural socialisation in French society is also demonstrated by the high divorce rate of children of Algerian descent born in France who marry a partner from Algeria (Tribalat, 1995, p. 85), a phenomenon which is also documented for North Africans and Turks in Western

9Portuguese in the Netherlands established their own parishes, because they did not feel at home in Dutch Catholic churches. Not only because of the language barrier but also because their version of Catholicism differed from the Dutch traditions:Laarman (2007).

10Unfortunately the rates on Belgium (in 1991) are a mixture of the first and second generation (see Table 4 in the Appendix A).

Table 7

Intermarriage rates for women of the second generation (1992–2003)

Netherlands (2003) Germany (2000) France (1992) UK (2002)

Tu Mo Wi In Tu It Gr Yu Al Sp Po Pa Ba Wi In

GW-NER 5 7 6 15

COL-NER 2 12 15

GW-ER 30 19 28 65 47

COL-ER 62^a 29

Key: SeeTable 1. Data for Belgium are not available for the second generation.

Source: see Appendix A, Table 1 (only the rates from 2000), 6 (2003), 12 (2002), 13 (1992).

a Average of Surinamese (44%) and Antilleans (79%).

(11)

Europe (Huis & Steenhof, 2003, p. 35–37; Neyrand &

M'Sili). Although at first sight such French–Algerian couples make a good cultural fit, very soon the differences in socialisation make themselves felt and many marriages are broken up within a few years. This shows that the second generation born in Western Europe has become much more European than many realize.

The pattern found for Algerians diverges strongly from that of Moroccans and Turks in countries like Belgium, the Netherlands and Germany. According to Todd in his discussion of Turks in Germany, this is first of all explained by the negative image that German nationals have of Turks and therefore their reluctance to marry them, and secondly by the Islamisation of Turks in Germany and the consolidation of the endogamous family system (Todd, 1994, p. 168–169, 178, 182). Other researchers have added, also with respect to Turks in the Netherlands, that most Turkish parents loathe the idea of their children marrying natives and are even critical of (second generation) partners within the Turkish group who are born in the receiving society. Instead, many parents prefer a Turkish born daughter or son-in-law, who is believed to be more trustworthy and traditional (Böcker, 1994, 2000). This pressure to choose partners from the country of origin, which we also found among Moroccan migrants, is further stimulated by the unintended effects of restrictive aliens policies. When other immigration channels were closed in the mid 1970s, marriage migration was one of the few options left. This means that there is a strong social pressure within Turkish and Moroccan groups for children born in Europe to marry a partner from the country of origin of their parents (Nelissen & Buijs, 2000, p. 190). In the Netherlands for example, almost three quarters of the Turkish youth who married between 1990 and 1995, were pressured by their parents to choose a partner from Turkey (Böcker, 2000, p. 164–165).

From this, however, we should not draw the conclusion that the second generation blindly follows the preferences of the parents, nor that this will automatically lead to lower social mobility. Moreover, we should realize that the choice of a partner from the country of origin is deeply gendered. As was evident in Tables 4–7, these communities have higher rates of intermarriage for men than for women, which is often explained by the lack of agency of women in choosing their future spouse: a feature that is recounted over and over again in the debates on the position of women in Islam (although it is surely not exclusively Islamic). But agency is not the only explanation. Dutch research has shown that Moroccan and Turkish women who decide to marry within their own ethnic and religious group,

often prefer a partner with a similar educational background. As many of these women are on average better educated than their male counterparts, they have difficulties finding a suitable partner in Europe and look for a groom in Turkey or Morocco. Moreover, these (higher educated) men in general have cultural views that fit better with the emancipated norms and aspira- tions that the Europe born women have developed. Men from these groups, on the other hand, on average have very low educational qualifications and hold conven- tional ideas on gender roles. Therefore they follow the opposite path by looking for a bride from the country of origin of their parents from a traditional background and no or very limited education (Hooghiemstra, 2003). The same holds true for Hindustani migrants from Surinam in the Netherlands (de Koning & Bartels, 2004).

In some debates on integration, endogamous marriages are seen as proof of the backwardness of migrant cultures and their unwillingness to integrate. Mixed marriages are thus put forward as modern and in-group marriages as traditional. This dichotomy dominates the debates, but is in reality less straightforward. Research has shown that in recent years the number of marriages of Western European men and migrant women from the Philippines and Thailand has exponentially risen. These men state that they preferred a spouse from these countries for they consider them more obedient, caring and traditional than Western European women (de Hart, 2004).

The third peculiarity of the data presented inTables 4–7 pertains to the colonial background in general.

When we link the intermarriage patterns of colonial migrants to the typology of family systems as elaborated by Todd, at first sight the picture seems to be clear. In cases where the family system of colonial migrants was matriarchal, nuclear and exogamous (Todd, 1994, 344– 345), as in the case of the descendants of African slaves in the Caribbean who went to France, Great Britain and the Netherlands (their metropolis), this premigration legacy lead to high intermarriage rates, notwithstanding racist stereotypes (van Amersfoort & van Niekerk, 2006). With other words, the prevalent (exogamous) family system (with insignificant numbers of cousins marrying each other) makes a better fit with the European egalitarian nuclear family and reinforces the effects of colonial socialisation. Within this family system, marriage is not the norm and many partners cohabitate. This indicates that the data presented inTables 4–7 do not tell the whole story. The actual number of mixed liaisons will probably be higher. Existing colonial racist hierarchies strength- ened the tendency for West Indians in these three countries to cross the racial boundary (‘colouring up’),

(12)

because a light skin colour meant a higher social status (Foner, 1998; van Niekerk, 2002). Muslim and Hindu migrants from former colonies in the Netherlands, France and the UK do not follow the same pattern, as religion dovetailed with endogamous and patrilineal family systems.

Finally, and fourthly, recent research by Raya Mutarrak on intermarriage patterns in Great Britain, problematises the sweeping assumption that increased intermarriage rates coincides with upward social mobility, and thereby structural integration. Thus, among the second generation of ‘Blacks’, both from the Caribbean and other origins, increased intermarriage rates do not necessarily correlate with higher educational attainments. Both men and women who are highly educated do not intermarry more and to some extent even less. Mutarrak explains this by pointing at the segmented nature of the integration process:

“Another interesting pattern is that Black African &

Black Other men with low qualifications and women with no qualifications have a higher propensity to marry a white partner than their counterparts with high qualifications. High qualifications do not seem to be the main driver of interethnic marriage for Black Caribbeans either. This implies that statisti- cally having no or low qualifications is a determi- nant of intermarriage for Black Caribbean and Black African & Black Other. In fact, Blacks are found to be relatively well-integrated in British society but likely to be associated with the white working class community rather that the service class (Peach, 1996b). This explains why high educational qualifications are not so important in promoting intermarriage in the case of Blacks because they have already been well-integrated segmentedly into the working class society.” (Mutarrak, 2007, p. 59) In fact, Mutarrak points at a general sociological phenomenon linked to social mobility. Blacks who rise on the social ladder thereby‘leave’ the white group they know best and enter a segment of British society which is relatively uncharted terrain. For them, as for whites with a working class background, it will take time to find partners with the same cultural capital, to borrow from Bourdieu, and this may increase the chance of marrying blacks who went through the same experience.

6. Conclusion

The migration history of Western Europe since the middle of the 19th century, which has long been neglected, shows that the slow and often painful

intergenerational integration process of post war migrants is nothing new. Irish newcomers in the United Kingdom and Italians in France for example, had to overcome major barriers in order to become integrated in the receiving societies. Apart from restricted social mobility and residential segregation, high rates of ethnic endogamy in the first and partly the second generation testify to this. When we compare these with intermarriage rates of the children of post-war guest workers and colonial migrants in Western European countries, the topic of this paper, we see some remarkable differences. Whereas most Southern Europeans show a similar pattern as their 19th century predecessors, this is less the case with the descendants of colonial and labour migrants (guest workers) whose roots lie outside of Europe.

To understand this difference we have compared the marriage patterns of both European and non-European immigrants in France, Germany, the United Kingdom, Belgium and the Netherlands in the post World War II period. The first conclusion is that religion is an important variable. Migrants whose faith has no tradition in Western Europe, intermarry much less than those whose religious backgrounds correspond with the religion that is common in the country of settlement.

Ethnic endogamous marriages in Western Europe are most conspicuous in Hindu and Muslim communities, regardless of their migratory background (guest workers or colonial migrants). Not only religious or theological restrictions on intermarriage play a role here. Endogamous, patrilineal family systems and ethno-national identification influence the propensity to intermarry as well. Muslim migrants seldom marry non Muslims, neither do they normally cross ethnic boundaries; in other words Moroccans do not marry Turks and vice versa. Even when the religion of migrants does not deviate, this not automatically promotes intermarriage, at least not in the first generation, as is illustrated by Portuguese migrants in France, an almost homogeneously Catholic country. In this case the strong cultural ties loosen up, however, already in the second generation.

Whereas differences in religion diminish the propensity to intermarry, colour or ‘racial’ differences on the other hand seem to be less important. This goes counter what one would expect on the basis of the dominant literature on the United States. The European situation shows that it makes a world of difference whether a country has an internal or an external slavery tradition. The late abolition of slavery in the U.S. South and the subsequent discriminatory Jim Crow segregation laws, which were eliminated only a century later, created a high awareness of race and colour,

(13)

with the “one drop of blood” rule as its most extreme expression (Alba & Nee, 2003, p. 133). This racist legacy which exerts its influence until today, blocked mutual identification of black and white Americans and, among other things, produced a very low intermarriage rate.

The external slavery tradition and hence colonialism, of the Netherlands, France and the United Kingdom— especially in the Caribbean—also produced racist structures and a high colour awareness, but this was predominantly restricted to colony and was unable to make a fundamental imprint in the mother country.

Moreover, colonialism also had an important socialising influence. Many subjects in the colonies were converted to Christianity, learned the language of the colonizer and were introduced to European political systems and values (Coleman, 1994, p. 118). Thus colonialism reduced the social and cultural distance between migrants from former colonies and natives in the‘motherland’. Due to this premigration socialisation, colonial migrants were already familiar with key aspects of the society they migrated to.

Furthermore, in many cases they were often defined as full citizens of the ‘mother country’ (Hansen, 2000), which made identification by the native population easier. After an initial phase of negative and racist stereotyping, as in the case of the West Indians in the 1950s through the 1970s, these dark skinned colonial migrants became relatively quickly included in the ‘imagined community’ to use Benedict Anderson's term (Anderson, 1983). Among other things this resulted in a sharp rise of the intermarriage rates.

An exception to this rule were those colonial migrants who had retained their original, non European, religion, such as the Islamic Algerians and the offspring of Indian Hindu migrants in Suriname and the United Kingdom. The Algerians who settled in France, however, pose an interesting intermediary case. Although Muslim, their children are much more exogamous than Moroccans and Turks in other Western European states without colonial ties with North Africa, which is explained by the colonial relation with France and the ensuing cultural and political socialisation.

Finally, our analysis has shown that apart from religion and colonial links, the low propensity to intermarry among Turks and Moroccans may also have been influenced by the unintended effects of restrictive migration policies.

The increasingly stringent measures to complicate and hinder the immigration from Morocco and Turkey in countries like Germany, Belgium and the Netherlands from the mid 1970s onwards, made marriage migration into one of the few remaining legal channels. As a result, the pressure from the homeland on the second generation to marry someone from the country of origin of the parents increased and lead to a large inflow of marriage migrants,

which—at least for the Netherlands—has only started to decline in the beginning of the 21st century.

What the future holds is of course uncertain, but it may be expected that also among Muslims and Hindus intermarriage rates will rise slowly, caused by the weakening of non Western family systems, the ongoing upward social mobility, at least for a part of these groups and the diminishing importance of ethno-national ties.

6.1. Theoretical implications

Finally a short word on the implications of this paper for the modern assimilation theory, which serves as an overarching framework for both contact and barrier theories with respect to intermarriage. Although in general this heuristic model does a good job in explaining diverse outcomes of settlement processes in general and intermarriage rates in particular, it has its limitations due to the strong focus on the U.S. context. Most clearly this is expounded by the analytical exclusion of African Americans from most studies on the assimilation process of migrants in U.S. cities. Although the Great Migration brought millions of Southern blacks to Northern (and Western) cities from the First World War onwards (Gregory, 2005), where they competed with European migrants, their experiences and settlement processes often play only a marginal role in the analysis. Their position in American society is, often implicitly, considered as hors concours, which confirms the broadly shared assumption that race, at least with African roots, is an almost unbridgeable barrier. The comparison with Europe, however, shows differently.¹¹

Furthermore the discussion of European intermarriage patterns in the 19th and 20th century showed that the attention to institutions, as rightly advocated by Richard Alba and Victor Nee, needs a more refined and layered elaboration. Institutions, often as barriers to intermarriage, do not only emanate from the receiving society, but also—be it less formalized—within migrant communities. Especially religions and family systems, but also organized nationalist feelings, can have a profound influence on how migrants think about endogamy. Finally, strong pressures to assimilate, often through institutionalized forms of discrimination and stigmatization, not only produce isolation and frustrate assimilation (with resulting low intermarriage rates), but can also stimulate assimilation by ‘passing’

11 For an exception to the isolationist tradition, see:Foner (2005);

andFoner and Fredrickson (2004).

(14)

mechanisms. These factors, together with a more comparative perspective, are not completely ignored in the new assimilation theory, but—as this study of Western European intermarriage patterns stresses— deserve to be included more systematically in historical and social scientist analyses.

Appendix A. Data on ethnic intermarriage in Western Europe 1980–2006

A.1. Germany

Table 2

Type of partner by sex and country of origin (1984–2002)

Turks Italians Yugoslavs Greeks Spanish

M F M F M F M F M F

7 2 24 11 22 28 18 5 21 15

Source:González-Ferrer, 2006, p. 175.

Original data source: GSOEP (1984–2002) (N=6000 households of which 1400 headed by non-Germans, due to oversampling of foreigners).

Table 5

Intermarriage rates of Turks, Moroccans, Antilleans and Surinamese in the Netherlands (1988–2002)

M F M + F

Turks 4

Moroccans 6

Antilleans 48 40

Surinamese 22 26

- Hindu 10

- Javanese 20

- Creole 26

Source:Kalmijn and Tubergen (2006, p. 385).

Original data source: SPVA data 1988–2002 (N=around 4000 households of Turks, Moroccans, Surinamese and Antilleans, consist- ing of some 8000 people. Response rates vary from 50–60%).

Table 1

Intermarriage rates of Turkish, Italian, Yugoslav and Greek nationals in Germany (1989–2000)

Turks Italians Yugoslavs Greeks

M F M F M F M F

1989 1st g. 6 1.5 37 8.5 18.5 16 17 7.5

2nd g. 8.5 9 60 51 66.5 57 26.5 9.5

1993 1st g. 6.5 2.5 40 15 15.5 18 17.5 8

2nd g. 4 3 69 24 23.5 32 16 5.5

1997 1st g. 9 3 39 15.5 19 16 16 8.5

2nd g. 7 4.5 45.5 23 27.5 18.5 30 11

2000 1st g. 13 7 37.5 13 21 18 17.5 12

2nd g. 13 5.5 51 30 30 28 35 19

Source:Schroedter (2006, p. 425–426).

Original data source: Microcensuses 1989, 1993, 1997 and 2000.

(N = 70% of 820,000 persons).

N.B. Immigrants and their descendants with a German passport are counted as‘German’.

Table 4

Intermarriage rates of Turks and Moroccans in Belgium in 1991

Turks Moroccans

M F M F

Western Europeans 5.6 1.8 16.8 6.1

Co-ethnics from Belgium 19.7 29.5 26.1 37.1 Co-ethnic from country of origin 74.7 68.7 57.1 56.8 Source:Lievens (1998, p. 123).

Original data source: Belgian Census 1991. (N = all married couples for which at least one partner had Turkish or Moroccan nationality. It concerns 11,174 Moroccan men, 7802 Moroccan women, 7378 Turkish men and 4934 Turkish women). Unfortu- nately no distinction is made between the first and second generation.

Table 3

Intermarriage rates for Turks (18–30 years old) in Germany (2000) Turks with German

nationality

Turks with Turkish nationality

M F M F

2000 11 2 6 2

Source:Haug (2004).

Original data source: Integrations Survey 2000. des Bundesinstituts für Bevölkerungsforschung (BiB) (N = 1200 Italians and Turks in the age of 18–30 and 1200 Germans as control group).

A.2 Belgium

A.3 Netherlands

Table 6

Origin of marriage partners of first and second generation migrants from Turkish, Moroccan, Surinamese and Antillean origin who married in 2003

With indigenous Dutch

With co-ethnic in The Netherlands

With co- ethnic from the country of origin

M F M F M F

Turks, 1st g. 7 9 25 29 61 57

Turks, 2nd g. 7 5 41 40 45 50

Moroccans, 1st g. 6 7 34 37 55 51

Moroccans, 2nd g. 14 7 49 51 30 36

Surinamese 1st g.^a 18 32 49 45 20 12

(continued on next page)

Immigration, intermarriage and the changing face of Europe in the post war period

post war period

Immigration, intermarriage and the changing face of Europe in the post war period

Leo Lucassen ⁎ , Charlotte Laarman

Leo Lucassen ^⁎ , Charlotte Laarman